of the Maritime University of Szczecin
Akademii Morskiej w Szczecinie
2017, 52 (124), 145‒153
ISSN 1733-8670 (Printed) Received: 19.08.2017
ISSN 2392-0378 (Online) Accepted: 10.11.2017
DOI: 10.17402/256 Published: 15.12.2017
GPS/GNSS spoofing and the real-time
single-antenna-based spoofing detection system
Evgeny Ochin
Maritime University of Szczecin, Navigation Faculty
1–2 Wały Chrobrego St., 70-500 Szczecin, Poland, e-mail: e.ochin@am.szczecin.pl
Key words: Global Navigation Satellite System, Global Positioning System, anti-terrorism, anti-spoofing,
transport safety, spoofer, spoofing detection algorithm
Abstract
The idea of C/A codes GPS/GNSS Spoofing (Substitution), or the ability to mislead a satellite navigation re-ceiver into establishing a position or time fix which is incorrect, has been gaining attention as spoofing has be-come more sophisticated. Various techniques have been proposed to detect if a receiver is being spoofed – with varying degrees of success and computational complexity. If the jammer signals are sufficiently plausible then the GNSS receiver may not realize it has been duped. There are various means of detecting spoofing activity and hence providing effective mitigation methods. In this paper, a novel signal processing method applicable to a single antenna handset receiver for spoofing detection has been described. Mathematical models and algo-rithms have been developed to solve the problems of satellite navigation safety. What has been considered in the paper is a spoofing detection algorithm based on the analysis of a civil satellite signal generated by mobile C/A GPS/GNSS single-antenna receivers. The work has also served to refine the civilian spoofing threat assess-ment by demonstrating the challenges involved in mounting a spoofing attack.
Introduction
Modern satellite navigation is based on the use of no-request range measurements between a naviga-tional satellite and the user. It means that the infor-mation about the satellite’s coordinates, given to the user, is included in the navigation signal. This meth-od of range measurement is based on the calculation of the receiving signal’s time delay compared with the signals generated by the user’s equipment.
Satellite based positioning provides the world’s most precise location information. It is possible to acquire positioning anywhere in the world that GNSS satellite signals are available, at any time of day, at data rates of up to 100 Hz. Measurements can be generated in real time or processed post-mission to achieve the highest level of accuracy.
GNSS technology is most frequently used to: • Determine the location of an object on, or with
respect to, the Earth for navigation;
• Locate an object with respect to another object for tracking purposes.
The positioning information typically provided includes a horizontal domain (latitude/longitude or easting/northing) and a vertical domain (height).
The definitions used in this article are:
1. GNSS – Global Navigation Satellite System {GLONASS: www.glonass-iac.ru, NAVSTAR GPS: www.navcen.uscg.gov, BEIDOU: en.bei-dou.gov.cn, GALILEO: www.gsc-europa.eu, QZSS: www.qzs.jp/en/services}.
2. Sati, i ,1N , N ≥ 4 – the navigation satellites as
the spacefaring component of GNSS. {In an ideal case, when the measurements are precise and sat-ellite time is identical to the user’s equipment time, the user’s position can be realized with only 3 sat-ellites. However the satellites time actually differs from the time on the user’s equipment. So, one more coordinate is necessary to find the user’s position – the time drift between the user’s equipment and
z
Sat1 Sat2 SatN
GNSS Antenna
x
Figure 1. GNSS: Sati – Satellites; i1,N, N ≥ 4; the visible
part of GNSS satellite constellation
the satellite time. That is why four satellites are needed to solve the navigation problem.}
3. Spoofing – an attack on a GNSS signal, in an attempt to deceive the GNSS receiver by transmit-ting powerful false signals that mimic the signals from the true GNSS, exceeding the power of these true signals.
4. Spoofer – complex computer and radio equip-ment for the impleequip-mentation of GNSS spoofing. 5. (x, y, z) – the real coordinates of a vehicle (victim).
If the vehicle is a 2D vehicle (ship, vessel, boat, car, etc.), the height coordinate (z) can be omitted, and the minimum number of navigation satellites required can be reduced to three (i ,1N, N ≥ 3). 6. (xv, yv, zv) – the precise coordinates of the vehicle.
7. (xˆv,yˆv,zˆv) – the calculated coordinates of the
vehicle using the GNSS.
8. (xs, ys, zs) – the precise coordinates of the
recep-tion antenna of the spoofer.
9. (xˆs,yˆs,zˆs) – the calculated coordinates of the
reception antenna of the spoofer.
10. We also denote for i ,1N, N ≥ 4 (if the vehi-cle is a 2D vehivehi-cle (ship, vessel, boat, car, etc.), the height coordinate (z) can be omitted and the minimum number of navigation satellites can be reduced to three (i ,1N, N ≥ 3)):
(xi, yi, zi) – the coordinates of Sati;
Tiv – the propagation time from Sati to the
vehi-cle in a vacuum;
– the propagation time from Sati to the
vehi-cle in the real atmosphere;
Div – the measurement result of the distance
from Sati to the vehicle (the vehicle’s
pseudo-range);
Dsv – the distance from the spoofer to the victim;
Δtsv – the signal transit time from the spoofer to
the victim;
Δρi – unknown error of the measurement result
of the distance from Sati to the vehicle. GNSS positioning
The distance from a vehicle (Figure 1) to the sat-ellites Sati, can be written as:
4 , ,1 2 2 2 N N i cT z z y y x x D v i v i v i v i v i (1) Since the measurement of the distance from the vehicle to the satellites is carried out by measuringthe propagation time v
i v i v i T T Tˆ of the GNSS signals from Sati to the vehicle, then (1) can be
rep-resented as (excluding time synchronization errors):
v i Tˆ
4 , ,1 ˆ 2 2 2 N N i T T c z z y y x x v i v i v i v i v i (2) Since Δρi = cΔTiv, then equation (2) can bewrit-ten in the form:
4 , ,1 ˆ 2 2 2 N N i T c z z y y x x v i i v i v i v i (3) The navigation processor in the vehicle solves the system of the equations (3), and calculates the posi-tion of the vehicle (xv, yv, zv) and the timing errors onboard Δt, which are then used to correct the GNSS navigation clock.
v v v
i v v v v v v v v v z y x z z y y x x z z y y x x z z y y x x , , 1,3 , Sat for algorithm Iteration 3 2 3 2 3 2 3 2 2 2 2 2 2 2 1 2 1 2 1 2 1 i (4) The iterative algorithm is based on the mech-anism of sequential reduction of the inaccurate (usually quartz) timer of the user, to the accurate (usually atomic) clocks onboard the navigation satellites (this article does not consider the timing errors, Δt).Because Δρi is not an unknown value, instead of
the exact value (xv, yv, zv) we will get an approximate
results of the measurements (xˆv,yˆv,zˆv):
v v v
i v v v v v v v v v z y x z z y y x x z z y y x x z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 2 3 2 3 2 3 2 2 2 2 2 2 2 1 2 1 2 1 i (5)Spoofing Antenna GNSS Antenna Spoofer Victim x
z Sat1 Sat2 Sat3
D1
Figure 3. GNSS Spoofer broadcasts a recorded signal of the visible part of the GNSS satellite constellation
Spoofing Antenna GNSS Antenna Spoofer Victim x
z Sat1 Sat2 Sat3
D1
z Sat1 Sat2 Sat3
Spoofer
x
The record of GNSS signals The broadcasts of the GNSS signals
GNSS spoofing
A spoofer transmits the simulated signals of sev-eral satellites; (Hartman, 1996; Humphreys et al., 2008; Humphreys et al., 2009; Rawnsley, 2011; Tippenhauer et al., 2011; Broumandan et al., 2012; Zaragoza & Zumalt, 2013). If the level of the sim-ulated signals exceeds the level of the signals from real satellites, the GNSS receiver captures the false signal and calculates the false coordinates.
We have distinguished the following spoofing modes:
A. A spoofer is motionless and broadcasts signals of the visible part of the GNSS satellite constella-tion, and then a repeater of the GNSS signals is used as the spoofer (Figure 2).
C. A spoofer is motionless and broadcasts a signal of the visible part of the GNSS satellite constel-lation with the introduction of signal delays from each of the GNSS satellites, and then a repeat of the GNSS signals with a programmed signal delay from each of the GNSS satellites is used as a spoofer (Figure 4).
Figure 4. GNSS Spoofer broadcasts a recorded signal of the visible part of the GNSS satellite constellation with the possi-bility of a programmed signal delay for each satellite
D. A spoofer is motionless and broadcasts a simulat-ed GNSS signals, and then a GNSS-signal simu-lator is used as a spoofer (Figure 5).
E. A spoofer is mobile and broadcasts signals of the visible part of GNSS satellite constellation, and then a GNSS signal repeater is used as the spoofer (Figure 6).
F. A spoofer is mobile and broadcasts a recorded signal of the visible part of the GNSS satellite constellation, and then the GNSS recorder is used as the spoofer (Figure 7).
G. A spoofer is mobile and broadcasts signals of the visible part of the GNSS satellite constellation
Figure 2. GNSS Spoofer broadcasts signals of the visible part of the GNSS satellite constellation
B. A spoofer is motionless and broadcasts a record-ed signal of the visible part of the GNSS satellite constellation, and then the GNSS recorder is used as the spoofer (Figure 3).
Spoofing Antenna GNSS Antenna Spoofer + programmer of signal delays Victim x
z Sat1 Sat2 Sat3
Spoofing Antenna GNSS simulator Victim x
z Sat1 Sat2 Sat3
D
Figure 5. GNSS Spoofer broadcasts simulated GNSS signals
Directional Antenna GNSS Antenna x z
Figure 6. Mobile GNSS Spoofer broadcasts signals of the visible part of the GNSS satellite constellation
Directional Antenna GNSS Antenna x z z x
The record of GNSS signals The broadcasts of the GNSS signals
Figure 7. Mobile GNSS Spoofer broadcasts a recorded signal of the visible part of the GNSS satellite constellation
Directional Antenna GNSS Antenna x z
Figure 8. Mobile GNSS Spoofer broadcasts a recorded signal of the visible part of the GNSS satellite constellation with the possibility of a programmed signal delay for each satellite
Directional Antenna
x z
Figure 9. Mobile GNSS Spoofer broadcasts a simulated GNSS signals
with the introduction of a signal delay from each of the satellites, and then a repeater of the GNSS signals with a programmed signal delay from each of the GNSS satellites is used as a spoofer (Figure 8).
H. A spoofer is mobile and broadcasts a simulated GNSS signals, and then a simulated GNSS-sig-nal is used as a spoofer (Figure 9).
In this article, only mode A has been considered.
In this mode a spoofer is motionless and broad-casts signals of the visible part of the GNSS satel-lite constellation, and then a repeater of the GNSS signals is used as the spoofer (Figure 2). A victim receives the same signal as the spoofer, but with some delay Δtsv. It means that all receivers in the
spoofing zone calculate the same false coordinates, regardless of the distance from the spoofer to the victim:
Spoofer
Spoofing detector
x
z Sat1 Sat2 Sat3
Y1
Y2
D1
D2
D1–2
Figure 10. The spoofer and the dual-antenna spoofing detector (SD): Y1 and Y2 – antennas of the SD; D1 and D2 –
the distances from the antenna of the spoofer to the anten-nas of the SD, D1–2 – the distance between the antennas of
the SD
s s s
i v s v v v v s v v v v s v v v z y x D z z y y x x D z z y y x x D z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 2 3 2 3 2 3 2 2 2 2 2 2 2 1 2 1 2 1 i (6) where Dsv = cΔtsv. Detection of spoofingFor the detection of GNSS spoofing, various methods have been suggested. We have listed some of them.
• Detection based on the determination of the direc-tion to the radiadirec-tion source of the spoofer, com-paring the phases of the signal to several antennas. • Detection based on the definition of the Doppler
frequency shift.
• Using the military GNSS signal as a reference (without the need to know the encryption key). • Comparing the indications of the inertial
naviga-tion system and the data from the GNSS receiver. Dual-antenna spoofing detector
The Spoofing Detector (SD) requires the use of two antennas (Figure 10) (Dobryakova, Lemiesze-wski & Ochin, 2012; 2013; 2014; Ochin et al., 2013; Dobryakova & Ochin, 2014; Psiaki et al., 2014). The distance between the antennas is denot-ed as D1–2.
Measuring the distance between antennas in normal navigation mode
The spoofing detector measures the coordinates of the antennas Y1 and Y2:
1 1 1
1,3 , Sat for algorithm Iteration 2 1 3 2 1 3 2 1 3 2 1 2 2 1 2 2 1 2 2 1 1 2 1 1 2 1 1 ˆ , ˆ , ˆ i v v v i v v v v v v v v v z y x z z y y x x z z y y x x z z y y x x (7) where (xv1, yv1, zv1) – denotes the unknown precisecoordinates of the antenna Y1, and (xˆv1,yˆv1,zˆv1) –
denotes the calculated coordinates of the antenna Y1.
2 2 2
1,3 , Sat for algorithm Iteration 2 2 3 2 2 3 2 2 3 2 2 2 2 2 2 2 2 2 2 2 1 2 2 1 2 2 1 ˆ , ˆ , ˆ i v v v i v v v v v v v v v z y x z z y y x x z z y y x x z z y y x x (8)where (xv2, yv2, zv2) – denotes the unknown precise
coordinates of the antenna Y2, and (xˆv2,yˆv2,zˆv2) –
denotes the calculated coordinates of the antenna Y2.
The measurement results differ by some unknown value and, accordingly, the distance estimate Dˆ12
between the antennas will be comparable with the magnitude of D1–2:
2 1 2 2 1 2 2 1 2 2 1 2 1 ˆ ˆ ˆ ˆ ˆ ˆ ˆ D z z y y x x D v v v v v v (9)Measurement of the spacing setween the antennas in spoofing mode
A victim receives the same signal as the spoof-er, but with some delay Δtsv. It means that all the
receivers in the spoofing zone calculate the same false coordinates, regardless of the distance from the spoofer to the victim:
s s s
i v s v v v v s v v v v s v v v z y x D z z y y x x D z z y y x x D z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 1 2 1 3 2 1 3 2 1 3 1 2 1 2 2 1 2 2 1 2 1 2 1 1 2 1 1 2 1 1 i (10)where Dsv1 = cΔtsv1 – the distance from the spoofer
to the antenna Y1, and (xˆs,yˆs,zˆs) – the calculated
coordinates of the spoofer using an antenna Y1.
s s s
i v s v v v v s v v v v s v v v z y x D z z y y x x D z z y y x x D z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 2 2 2 3 2 2 3 2 2 3 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 1 2 2 1 i (11) where Dsv2 = cΔtsv2 – the distance from the spooferto the antenna Y2, and (xˆs,yˆs,zˆs) – the calculated
coordinates of the spoofer using an antenna Y2.
In this case, the distance between the antennas Y1
and Y2 is defined as:
ˆ ˆ
ˆ ˆ
ˆ ˆ
0 ˆ 2 2 2 2 1 xsxs ysys zszs D (12) The Decisive Rule 1Comparing (9) and (12), the decisive rule for detecting spoofing can be written as:
then Spoofing else GNSS
ˆ if D1 2 D
(13) where D – discriminant, is determined on the basis of statistical studies at the design stage for a real detection system.
The algorithm for spoofing detection by estimating the dispersion of the pseudorange difference of two antennas
In the normal navigation mode, the pseudoranges of the antennas Y1 and Y2 differ from each other by
some unknown, but significantly different value:
i i
i
ˆ ˆ ˆ (14)
Therefore, the root-mean-square deviation (RMSD) in the differences in the pseudoranges of the antennas Y1 and Y2 will be significantly different
from zero:
0 1 ˆ ˆ 1 ˆ ˆ 1 2 1 2
NN N i i i N i i i gnss (15) In the spoofing mode, the pseudoranges of the antennas Y1 and Y2 differ from each other bya certain constant value equal to D1 – D2. In this case
the RMSD difference in the pseudoranges of anten-nas Y1 and Y2 is practically zero, that is:
0
s
(16)
The decisive rule 2
Comparing (15) and (16), the decisive rule for spoofing detection can be written as:
GNSS else Spoofing then 2 1 ˆ ˆ 1 ˆ ˆ if 2 1 1 2 s gnss N i i i N i i i NN (17) If we take σgnss >> σs, then the decisive spoofingdetection rule can be written as:
GNSS else Spoofing then 2 1 ˆ ˆ 1 ˆ ˆ if 2 1 1 2 gnss N i i i N i i i NN (18)Discussion of the decisive rules
The spoofing detector can be designed on the basis of one of the decisive rules or on the basis of any combination of decision rules. In any case, it is necessary to calculate the probabilities of the “False alarm (false positives)” and “Missing target (false negatives)” events (Table 1).
Table 1. Mistakes of a decision of the first kind (False alarm) and the second kind (Missing target)
The decisive rule or
combination of decision rules GNSSValid modeSPOOFING Solving
of Spoofing Detector
GNSS The solution
is right Missing target SPOOFING False alarm The solution
is right
The questions of optimal design and selection of boundary conditions with the aim of minimiz-ing the probabilities of “false alarm” and “missminimiz-ing target” are beyond the scope of this article. Here it should be noted that one of the widely used tech-niques is the application of Bayes theorem (or Bayesian formula) (Dobryakova, Lemieszewski & Ochin, 2012).
Single-antenna spoofing detector in case the vehicle is moving
Suppose that the vehicle is moving in an arbi-trary direction. On the spoofing detector we install a single-antenna Y (Figure 11). The position of the antenna at the time tʹ is denoted as Yʹ, the position of the antenna at the time tʺ = tʹ + Δt is denoted as Yʺ and the distance between the two antenna positions is denoted as D1–2.
Measuring the distance between two positions of single-antenna in normal navigation mode
The spoofing detector measures the coordinates of the antenna Y in two positions:
v v v
i v v v v v v v v v z y x z z y y x x z z y y x x z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 2 3 2 3 2 3 2 2 2 2 2 2 2 1 2 1 2 1 i (19)where (xv’, yv’, zv’) – the unknown precise coordinates
of the antenna Y at the time tʹ, and (xˆv,yˆv,zˆv) – the
calculated coordinates of the antenna Y at the time tʹ:
v v v
i v v v v v v v v v z y x z z y y x x z z y y x x z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 2 3 2 3 2 3 2 2 2 2 2 2 2 1 2 1 2 1 i (20)where (xvʺ, yvʺ, zvʺ) – the unknown precise
coordi-nates of the antenna Y at the time tʺ = tʹ + Δt, and, (xˆv,yˆv,zˆv) – the calculated coordinates of the
antenna Y at the time tʺ = tʹ + Δt.
The distance between the antenna Y at the time tʹ and the antenna Y at the time tʺ = tʹ + Δt will be comparable with the magnitude D1–2:
2 1 2 2 2 2 1 ˆ ˆ ˆ ˆ ˆ ˆ ˆ D z z y y x x D v v v v v v (21)Measurement of the spacing between two positions of a single-antenna in spoofing mode
A victim receives the same signal as the spoof-er, but with some delay Δtsv. It means that all the
receivers in the spoofing zone calculate the same false coordinates, regardless of the distance from the spoofer to the victim:
s s s
i v s v v v v s v v v v s v v v z y x D z z y y x x D z z y y x x D z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 2 3 2 3 2 3 2 2 2 2 2 2 2 1 2 1 2 1 i (22)where Dsvʹ = cΔtsvʹ – the distance from the spoofer
to the antenna Y at the time tʹ, and (xˆs,yˆs,zˆs) –
the calculated coordinates of the spoofer using the antenna Y at the time tʹ:
s s s
i v s v v v v s v v v v s v v v z y x D z z y y x x D z z y y x x D z z y y x x ˆ , ˆ , ˆ 1,3 , Sat for algorithm Iteration 2 3 2 3 2 3 2 2 2 2 2 2 2 1 2 1 2 1 i (23) where Dsvʺ=cΔtsvʺ – the distance from the spoofer tothe antenna Y at the time tʺ=tʹ+Δt, and (xˆs,yˆs,zˆs)
– the calculated coordinates of the spoofer using the antenna Y at the time tʺ = tʹ + Δt.
In this case, the distance between the antenna Y at the time tʹ and the antenna Y at the time tʺ = tʹ + Δt is defined as:
ˆ ˆ
ˆ ˆ
ˆ ˆ
0 ˆ 2 2 2 2 1 xsxs ysys zszs D (24) Spoofer xz Sat1 Sat2 Sat3
Y' Y"
D1
D2
D1–2
Figure 11. Spoofer and single-antenna spoofing detector (SD): Yʹ and Yʺ – two positions of single-antenna Y; D1 and
D2 – the distances from the spoofer’s antenna to the SD
antenna Y; D1–2 – the distance between two positions of
The decisive rule
Comparing (21) and (24), the decisive rule for detecting spoofing can be written as:
then Spoofing else GNSS
ˆ if D1 2 D
(25) where D – discriminant, is determined on the basis of statistical studies at the design stage of a real detection system.
Rotating single-antenna spoofing detector The antenna can be installed on a rotating turret, the radius of which is about one meter. In this case, the diameter of the antenna’s rotation circle is equal to two meters. The spoofing detector measures the coor-dinates of the antenna in two positions (Figure 12):
Figure 12. Rotating single-antenna spoofing detector, in the case when the vehicle is in motionless: Δt = T/2, where T – rotation period of antenna Y
The measurements are performed in accordance with (23–24), and the decisive rule in accordance with (25).
The equipment for experimental studies For the experimental studies, standard equipment was used, including two Holux GR-213u GNSS receivers (Figure 13).
This equipment allowed for experimental studies in the following four modes (Figure 14).
1. The ship is moving or not moving. Two fixed antennas are in two separate positions (https:// goo.gl/1Fk5Na).
2. The ship is moving or not moving. One antenna is used in series in two positions. After the first position measurement is taken, the antenna is transferred to another position and the second measurement is performed. Y at the time t' x y R Y at the time t" = t' + Δt
Figure 13. The equipment used for experimental studies: HOLUS – GNSS-receiver (Dobryakova, Lemieszewski & Ochin, 2014)
Y'(x', y') Y"(x", y") Y(x', y') at the time t' Y(x", y")
at the time t" = t' + Δt
Y(x', y') at the time t' Y(x", y") at the time t" = t' + Δt
Y at the time t'
Y at the time
t" = t' + Δt 1. Y'(x', y') and Y"(x", y") −
two positions of two antennas Y' and Y"
2. Y(x', y') and Y(x", y") −
two positions of single-antenna Y
3. Y(x', y') and Y(x", y") −
two positions of single-antenna Y 4. Rotating single-antenna: Δt = T/2, where T − rotation period of antenna Y Figure 14. The four modes used in the experimental studies
3. The ship is moving. One antenna is used in series in two positions. The second measurement is per-formed after a certain period of time Δt.
4. The ship is moving or not moving. One antenna is used in series in two positions (rotating sin-gle-antenna). The first measurement is performed with the angle of antenna’s rotation α, and the second measurement with the angle of antenna’s rotation α + 180°.
Comparing methods 2, 3 and 4, it can be noted that there is not a fundamental difference between these methods in terms of performing a sequence of measurements, using an algorithm to process the signals, and interpreting the results of the measurements.
Summary and conclusions
The risk of losing the GNSS signal is growing every day. The equipment necessary to manufacture GNSS “jamming” and/or “spoofing” systems are now widely available, and this type of attack cannot only be deployed by the military, but also by terror-ists. The distortion of the signal includes signal cap-ture and playback at the same frequency with a slight shift in time and with greater intensity, in order to deceive the electronic equipment of the victim and, therefore, the operator if there is one on board the vehicle. The price of one chipset for such equipment is in the range of 1–10 thousand Euros, depending on the dimensions and weight parameters.
It is important to emphasize that GNSS is not only used for navigation. In the framework of the current threat model, GNSS interference is needed in order to drown out the reference signal of syn-chronous time that is used in a distributed network of electronic radio devices. That is, GNSS allows you to synchronize with a very accurate time signal on stand-alone passive devices.
Synchronous time is also necessary for data trans-fer in a low-visibility communication mode: receiv-ers and repeatreceiv-ers must have a total time to correctly adjust the correlation parameters, allowing them to isolate the masked signal, which is indistinguish-able from noise for an outside observer. And this, of course, is not a complete list.
References
1. Broumandan, A., Jafarnia-Jahromi, A., Dehghanian, V., Nielsen, J. & Lachapelle, G. (2012) GNSS Spoofing
Detection in Handheld Receivers based on Signal Spatial Correlation. Position Location and Navigation Symposium (PLANS), April 23–26, 2012, Myrtle Beach, SC, USA. Available from: http://ieeexplore.ieee.org/abstract/docu-ment/6236917/ [Accessed: August 15, 2017]
2. Dobryakova, L. & Ochin, E. (2014) On the application of GNSS signal repeater as a spoofer. Scientific Journals of the Maritime University of Szczecin, Zeszyty Naukowe Akademii Morskiej w Szczecinie 40 (112), pp. 53–57.
3. Dobryakova, L., Lemieszewski, Ł. & Ochin, E. (2012) Antiterrorism – design and analysis of GNSS antispoofing algorithms. Scientific Journals of the Maritime University of Szczecin, Zeszyty Naukowe Akademii Morskiej w Szczecinie 30 (102), pp. 93–101.
4. Dobryakova, L., Lemieszewski, Ł. & Ochin, E. (2013) The analysis of the detecting algorithms of GNSS-spoofing. Scientific Journals of the Maritime University of Szczecin, Zeszyty Naukowe Akademii Morskiej w Szczecinie 36 (108) z. 2, pp. 30–36.
5. Dobryakova, L., Lemieszewski, Ł. & Ochin, E. (2014) Design and Analysis of Spoofing Detection Algorithms for GNSS Signals. Scientific Journals of the Maritime Univer-sity of Szczecin, Zeszyty Naukowe Akademii Morskiej w Szc-zecinie 40 (112), pp. 47–52.
6. Hartman, R.G. (1996) Spoofing Detection System for a Satellite Positioning System. U.S. Patent No. 5,557,284, Sept. 1996. Available from: https://www.google.ch/patents/ US5557284
7. Humphreys, T.E., Kintner, P.M., Jr., Psiaki, M.L., Ledvi-na, B.M. & O’Hanlon, B.W. (2009) Assessing the Spoof-ing Threat. GPS World 20 (1), pp. 28–38.
8. Humphreys, T.E., Ledvina, B.M., Psiaki, M.L., O’Hanlon, B.W. & Kintner, P.M. Jr. (2008) Assessing the Spoofing Threat: Development of a Portable GPS Civilian Spoofer. Preprint of the 2008 ION GNSS Conference Savannah, GA, September 16–19, 2008 https://radionavlab.ae.utexas.edu/ images/stories/files/papers/ion2008r01_for_distributionW. pdf
9. Ochin, E., Lemieszewski, Ł., Lusznikov, E. & Dobryako-va, L. (2013) The study of the spoofer’s some properties with help of GNSS signal repeater. Scientific Journals of the Maritime University of Szczecin, Zeszyty Naukowe Akademii Morskiej w Szczecinie 36 (108) z. 2, 159–165.
10. Psiaki, M.L., O’Hanlon, B.W., Powell, S.P., Bhatti, J.A., Humphreys, T.E. & Schofield, A. (2014) GNSS Spoofing Detection using Two-Antenna Differential Carrier Phase. GPS World 25, (11), pp. 36–44.
11. Rawnsley, A. (2011) Iran’s Alleged Drone Hack: Tough, but Possible. [Online] December 2011. Available from: http:// www.wired.com/dangerroom/2011/12/iran-drone-hack-gps/ [Accessed: August 15, 2017]
12. Tippenhauer, N.O., Pöpper, C., Rasmussen, K.B. & Čapkun, S. (2011) On the Requirements for Successful GPS Spoofing Attacks. Proceedings of the 18th ACM conference
on Computer and communications security CCS’11, Chi-cago, Illinois, USA, October 17–21, 2011. Available from: https://www.cs.ox.ac.uk/files/6489/gps.pdf [Accessed: Au-gust 15, 2017]
13. Zaragoza, S. & Zumalt, E. (2013) Spoofing a Superyacht at Sea. [Online] July 2013. Available from: http://www. utexas.edu/know/2013/07/30/spoofing-a-superyacht-at-sea/ [Accessed: August 15, 2017]